Conventional techniques for texture and material analysis for visual image data are generally performed along sub-regions or lines that align with the X or Y axes of the image. Unfortunately, aligning with the X or Y axes introduces many problems in the image analysis. For example, sharp edges in an image, such as seams, paper edges, or shadows, exist in the real world and tend to distort the underlying image data of the image. Consequently, when texture and material analysis is performed along the X or Y axes of the image on image data that includes sharp edges, the result may not be reliable, as the sharp edges tend to not show up at all or to show up in a much greater degree than they should.
In addition, when the viewing angle of the image changes, the sharp edges cause the image data to change. The relationship between the sharp edges and the viewing angle changes also changes under these circumstances. This also leads to inaccurate results when performing image analysis on the underlying image data of the image.